The present proposal is designed to examine the effects of a new medication, Acamprosate on two central features of alcohol dependence: withdrawal and loss of control over drinking. Acamprosate has been approved as a treatment for alcohol dependence in several European countries based on evidence indicating a significant improvement in continuous abstinence rates following acamprosate treatment. However, although acamprosate attenuates alcohol drinking in rodents there is minimal information regarding alcohol drinking in humans. Subjects will be randomly assigned to receive either one of two doses (1500 or 3000 mg/day) of acamprosate, or placebo for 13 days. During the first seven outpatient days drinking behavior will be monitored. This will be followed by a four-day inpatient period when withdrawal will be monitored, followed by a fifth inpatient day when subjects will be exposed to an alcohol self-administration paradigm, during which deprivation-induced alcohol consumption behavior will be monitored. The specific aims of this proposal are: 1) To examine whether pretreatment with two doses of acamprosate, or placebo, for seven days prior to abstinence initiation attenuates the intensity of acute alcohol withdrawal. We hypothesize that acamprosate will dose dependently reduce the intensity of withdrawal-induced changes in neuroendocrine, physiological, subjective and cognitive responses; 2) To determine whether acamprosate attenuates responses to a fixed low dose of alcohol (0.03 gm/kg) following a period of alcohol deprivation. We hypothesize that acamprosate will reduce the stimulatory effects of, and craving for, alcohol; 3) To determine whether acamprosate dose dependently reduces alcohol self-administration following an initial priming dose. Better characterization of these effects will be important for determining the optimal sequencing and timing of pharmacotherapies such as acamprosate for alcohol dependence, and for increasing overall success rates of achieving and maintaining abstinence.